Optimal Transportation Plans and Portfolios for Synchromodal Container Networks

(1)

Erasmus University Rotterdam (EUR) Erasmus Research Institute of Management Mandeville (T) Building

Burgemeester Oudlaan 50

3062 PA Rotterdam, The Netherlands P.O. Box 1738

3000 DR Rotterdam, The Netherlands T +31 10 408 1182

E info@erim.eur.nl W www.erim.eur.nl

This dissertation proposes an integrated approach for optimising synchromodal container transportation, motivated by two separate trends in the container transportation practice in North-West Europe. On the one hand, competition in hinterland transportation and the societal need for a modal shift towards sustainable modes require more integrated network optimisation of container transports. On the other hand, hinterland users increasingly require a cost-eff ective, but fl exible and reliable delivery service. The concept of synchromodality was developed as an answer to these developments, combining effi cient planning with a business model based on customer-oriented transportation services. This dissertation contributes by bringing together optimal transport planning in intermodal networks and the design of an optimal fare class mix of customer-oriented services. It includes 5 new models for operating such a synchromodal transportation network: service network design, disturbance analysis, real-time decision support and two variants of the cargo fare class mix design. All models are developed with the perspective of a centralised operator in an intermodal container network, with scheduled services between a deep-sea terminal and multiple inland ports. These scheduled services can be trains or barges, but not necessarily both have to be available. All 5 models have been applied to case studies based on the intermodal container network of European Gateway Services (EGS), a subsidiary of Hutchison Ports ECT Rotterdam (ECT). About the author

Bart van Riessen obtained a Master degree in Mechanical Engineering from TU Delft and a Master degree in Econometrics from EUR. Afterwards, he started in a part-time position at ECT on hinterland developments and in a separate position at the Econometric Institute (Erasmus University Rotterdam) for his Ph.D. research, co-supervised by the Dept. of Maritime and Transport Technology (TU Delft). His aim is to bridge the gap between academic transportation research and the transportation and logistics industry.

The Erasmus Research Institute of Management (ERIM) is the Research School (Onderzoekschool) in the fi eld of management of the Erasmus University Rotterdam. The founding participants of ERIM are the Rotterdam School of Management (RSM), and the Erasmus School of Economics (ESE). ERIM was founded in 1999 and is offi cially accredited by the Royal Netherlands Academy of Arts and Sciences (KNAW). The research undertaken by ERIM is focused on the management of the fi rm in its environment, its intra- and interfi rm relations, and its business processes in their interdependent connections.

The objective of ERIM is to carry out fi rst rate research in management, and to off er an advanced doctoral programme in Research in Management. Within ERIM, over three hundred senior researchers and PhD candidates are active in the diff erent research programmes. From a variety of academic backgrounds and expertises, the ERIM community is united in striving for excellence and working at the forefront of creating new business knowledge.

ERIM PhD Series

Research in Management

448 BAR T V AN RIESSEN - Optimal T

ransportation Plans and Portfolios for Synchr

omodal Container Networks

Optimal Transportation

Plans and Portfolios for

Synchromodal Container

Networks

BART VAN RIESSEN

(2)

Optimal Transportation Plans and Portfolios for

Synchromodal Container Networks

Bart van Riessen

Erasmus University Rotterdam

Delft University of Technology

(3)

Cover illustration by B. van Ooijen (Rotterdam, 2018)

The research was partially supported by the NWO/STW VENI project ‘Intelligent multi-agent control for flexible coordination of transport hubs’ (project 11210) of the Dutch Technology Foundation STW, by SmartPort@Erasmus and by the Erasmus Center for Maritime Economics & Logistics. Chapter 6 was partially supported by NWO project Integrated Synchromodal Transport System Analysis (ISOLA, project no. 438-13-214).

(4)

Optimal Transportation Plans and Portfolios for

Synchromodal Container Networks

Optimale transportplanning en portfolio’s voor

synchromodale containernetwerken

Thesis

to obtain the degree of Doctor from the Erasmus University Rotterdam

by command of the rector magnificus Prof.dr. H.A.P. Pols

and in accordance with the decision of the Doctorate Board. The public defence shall be held on

Thursday the 22nd of March, 2018 at 13.30 hrs by

BART VAN RIESSEN born in Rotterdam

(5)

Doctoral Committee

Doctoral dissertation supervisors: Prof.dr.ir R. Dekker Prof.dr. R.R. Negenborn Other members: Prof.dr. A.W. Veenstra

Prof.dr. I.F.A. Vis Prof.dr. R.A. Zuidwijk

Erasmus Research Institute of Management – ERIM

The joint research institute of the Rotterdam School of Management (RSM) and the Erasmus School of Economics (ESE) at the Erasmus University Rotterdam Internet: http://www.erim.eur.nl

ERIM Electronic Series Portal: http://repub.eur.nl/ ERIM PhD Series in Research in Management, 448 ERIM reference number: EPS-2018-448-LIS

This publication (cover and interior) is printed by Tuijtel on recycled paper, BalanceSilk® The ink used is produced from renewable resources and alcohol free fountain solution.

Certifications for the paper and the printing production process: Recycle, EU Ecolabel, FSC®, ISO14001. More info: www.tuijtel.com

This thesis is also printed as:

TRAIL Thesis Series no. T2018/2, the Netherlands Research School TRAIL ISBN: 978-90-5584-232-2

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the author.

(6)

Dedicated to my grandparents, who have each in their very own way contributed to who I am now.

(7)

(8)

vii

Acknowledgements

The process to complete this dissertation has been a very interesting and challenging period in my life. I am grateful for the support and effort of my promotor Rommert Dekker who step-by-step enticed me into pursuing this PhD research. Also, I am grateful for the continuous support and meticulous comments by my other promotor Rudy Negenborn (TU Delft) that helped me to achieve the required level of quality, first during the work on my Master thesis and subsequently for each chapter in this dissertation. It is satisfying to me that he recently obtained his full professorship and is now acknowledged as my second promotor.

Also, my situation with a position in both Europe Container Terminals (ECT) and Erasmus University Rotterdam was only possible due to the support, confidence and persistence of Rommert Dekker and Paul Ham, my manager at ECT. I know that both of them are proud themselves on making this research possible: they successfully battled the issues created by rigid regulatory frameworks that are focused on full time PhD students or full time employees. Not only was this crucial for the practicality of the problems and insights in this dissertation, of course I liked the ‘special’ position they created for me as well.

I would like to thank prof Rob Zuidwijk, prof Albert Veenstra and prof Iris Vis for being part of my inner committee and their valuable evaluation and comments on this dissertation. Considering your expertise on these topics in academia and practice and the level of detail in all of your evaluations, I’m looking forward to our discussions during the defence and after that. Specifically, I would like to thank Albert for bringing me in contact with ECT in 2012 for my Master thesis. I appreciate that prof Lóránt Tavasszy and prof Bert de Groot agreed to participate in the opposition during my defense.

I am grateful to all my colleagues at ECT for trusting in me and for providing me with the opportunity to conduct research into the EGS network: first during my internship and subsequently for participating in the research for this dissertation. Also, I would like to thank the staff and fellow PhD students at the Econometric Institute, who where very supporting and helped me getting through all these years. Judith, I want to thank you for always answering my questions and involving me in social occasions, even though I was not often available at the campus. I apologise for not mentioning all colleagues that supported me over the course of

(9)

viii Optimal Transportation Plans and Portfolios for Synchromodal Container Networks

the years, but I would like to make a special mention of the following persons for their direct contributions: Albert, Amanda, Andy, Arno, Charlie, Charlotte, Elias, Erik, Fred, Geert, Hans, Ian, Jan, Jasper, Jean-Pierre, Jenda, Judith, Kevin, Kevin, Kim, Leo, Marco, Mark, Michel, Michiel, Mirjam, Panos, Paul, Paul, Rana, Raymon, Rob, Rob, Rutger, Sander, Sicco, Wando, Willemien, Yingqian and Yiyang.

I want to thank my close friends, for always enduring my talking about containers, Bart, Bram, Jan Willem, Jochem, Mehdi, Steven, Thomas and Willem. I can’t promise it won’t happen again.

Finally, I want to thank my closest family. My parents, Jaap en Carli, you have always given me your unconditional support and you have always shown me what is most important in life. Thank you so much. Tim, when we were kids, we could fight and play on the same day. I’m proud of our bond. Since we are adults you are my friend, brother, pupil, mentor and much more. Before I knew what a paranimph was, I knew you were it for me. Amanda, I didn’t go to the Maasvlakte expecting to find you, but I’m so happy I did. I’m glad we are not colleagues anymore, even though we so often seem to follow similar interests in our professional lives. Much more I want to thank you for what you brought to my personal life and I am looking forward to our future. I couldn’t be happier having you by my side, as a paranimph during the defence, and beyond.

Bart van Riessen,

(10)

ix

Clarification of contribution

The research for this dissertation was carried out at Erasmus University Rotterdam (EUR) and Delft University of Technology (TU Delft) in cooperation with business partner Hutchison Ports ECT Rotterdam (ECT), and its subsidiary European Gateway Services (EGS). The groundwork was laid during my internship at ECT for my master thesis research in 2012-2013, under supervision of my current promotors, prof Rommert Dekker (EUR) and prof Rudy Negenborn (TU Delft); and prof Gabriel Lodewijks (TU Delft). From 2013 onwards, I have worked in an industry position at ECT, in combination with a separate, independent position at Erasmus University Rotterdam for academic research in continued cooperation with my promotors. My industry position provided three crucial aspects for this dissertation: first-hand insight in the problems and developments at EGS, access to data for case studies and input on related topics from colleagues and from interns for whom I was company supervisor. Although none of my colleagues or interns provided direct input for this thesis, they provided crucial information and inspiration for the research.

The chapters of this dissertation are based on separate article publications and can be read independently. Therefore, each chapter has a separate introduction, literature overview and notational framework. The chapters are provided in chronological order of the underlying research. Consequently, each chapter contains some repetition on the topics of the EGS case and the concept of synchromodal container transportation networks. Here, for each chapter a clarification is provided regarding the contributing authors and the relation with ECT and/or EGS.

 Chapter 1: This introductory chapter is based on the research proposal, written by myself, under supervision of my promotors. For this dissertation it was extended with a recent literature overview.

 Chapter 2: This work is based on Part I of my master thesis. For this dissertation, the research was substantially extended mostly by myself, especially the case study in Section 2.5, under supervision of my promotors. The data used in the case study is directly obtained from EGS, but perturbed by myself before using in the research to protect the

(13)

xii Optimal Transportation Plans and Portfolios for Synchromodal Container Networks

confidentiality.

 Chapter 3: The work in this chapter is based on Part II of my master thesis. For this dissertation, the case study in the original study was extended mostly by myself, under supervision of my promotors. The data used in the case study is directly obtained from EGS, but perturbed by myself before using in the research to protect the confidentiality.

 Chapter 4: The majority of the research and writing for this chapter was done independently by myself, under supervision of my promotors. The studied operational planning decisions and decision support for the intermodal problem were identified by me and discussed with experts from EGS. Michel van de Velden contributed with a detailed review of the initial manuscript.

 Chapter 5: The majority of the research and writing for this chapter was done independently by myself, under supervision of my promotors. This research starts from the premise of a differentiated transportation portfolio offered by EGS. That premise is the direct result of the work by interns at ECT; Michiel Verkaik, Yiyang Lin and Jorge Lecona as well as one of my colleagues at ECT; Elias Becker. Kevin Wardana and Ishara Schutte verified the results independently in their bachelor theses.

 Chapter 6: This chapter was the result of cooperation with Judith Mulder, who contributed numerous times in brainstorming about the methodology, and its proofs. She provided the Proof 6.1. The remainder of the research, implementation and writing was done independently by myself, under supervision of my promotors.

 Chapter 7: The concluding chapter was written by myself to cover academic conclusions and practical implications, under supervision of my promotors.

None of the research was commissioned by ECT or EGS. ECT and EGS are acknowledged for providing me with the opportunity to conduct research into the EGS network during an internship and for participating in the research for this dissertation with me and my co-authors holding independent positions at Erasmus School of Economics and Delft University of Technology.

(14)

1 Introduction

In this introductory chapter the background, the problem statement and the outline of the research is described. We provide an overview of relevant research around topics of synchromodal container transportation. Synchromodality refers to creating the most efficient and sustainable transportation plan for all orders in an entire network of different modes and routes, by using the available flexibility. We identify three topics that are relevant for practical implementation of synchromodality. For each topic we describe practical relevance and introduce our research on these topics in the next chapters. Also, our case of European Gateway Services is introduced, a major network orchestrator of container transportation in the Rotterdam hinterland. This chapter is structured as follows. In Section 1.1, the problem statement is provided with three subtopics. In Section 1.2 the case is introduced that we use for illustrating our research results. In Section 1.3, an overview of literature is provided for synchromodal transportation in general (Section 1.3.1) and for each of the three considered topics (Section 1.3.2-1.3.4). Section 1.4 gives our research approach for the three topics. Section 1.5 provides an outline of the remainder of the dissertation.1

1.1 Background and problem statement

In recent years intermodal networks have received renewed attention for two reasons: focus on shifting containers from truck transportation towards barge or rail transportation and an increased competition on hinterland transportation between players in maritime transportation, especially in North West Europe. In the 900km Hamburg – Le Havre range, multiple major container ports are located. Port authorities have put focus on modal shift towards more environmental friendly transportation modes. E.g. the ports of Rotterdam, Antwerp and Hamburg have

1_{This chapter is an adapted version of Van Riessen, B., Negenborn, R. R. and Dekker, R.} (2015, September). Synchromodal Container Transportation: An Overview of Current Topics and Research Opportunities. In Proceedings of the 6th International Conference on Computational Logistics (ICCL'15), Delft, The Netherlands, (pp. 386-397). The original publication is available via http://dx.doi.org/10.1007/978-3-319-24264-4_27.

(15)

2 Optimal Transportation Plans and Portfolios for Synchromodal Container Networks

stated modal split requirements for the hinterland transportation of containers (Van den Berg and De Langen, 2014). In Port Vision 2030 the Port of Rotterdam Authority (2011) aims for a modal shift in the hinterland transportation of containers. In 2015, 53% of the containers were transported by truck between the terminals in the Port of Rotterdam and inland destinations in North-West Europe (Topsector Logistiek, 2016). In 2035 this must be reduced to 35%. To achieve this modal shift, containers must be transported on intermodal corridors, using barge or rail services between deep-sea terminals and inland terminals. Although intermodal corridors are operating already for decades, many practical problems remain: demand for container transportation is volatile, seasonal and imbalanced, resulting in low utilisations of rail and barge services. Also, because of the complexity of operations, and dependency on terminal infrastructure, a barge or train is sensitive to disruptions, resulting in late delivery. Several researchers have stressed the complexity of achieving the required modal shift, i.e. in Veenstra et al. (2012) the need for an integrated network approach is emphasised, and Van der Horst and De Langen (2010) mention the mind shift that is required for achieving more integrated inland transportation. Efficient planning methods for transportation are essential to achieve this, while meeting customer requirements for synchronising the container supply chain and a further reduction of delivery time, costs and emissions. These trends motivate the use of inland container transportation networks, with multiple possible transport corridors and modes. In such an intermodal network, containers can be transported by one or more consecutive rail and barge services, using

intermediate transfers of the containers at network terminals. This type of operation

potentially allows more balanced planning for higher utilisations and can offer alternative services in case of disruptions. However, it requires new methods to guarantee efficient operation, in terms of cost, reliability and emissions. These intermodal container transportation networks are generally formed by the cooperation of multiple barge service operators, rail service operators and terminals. Roso et al. (2009) defined the concept of a dry port: ‘a hinterland terminal in close connection to the sea port, where customers can leave or pick up their standardised units as if directly at a sea port.’ Based on this concept, Veenstra et al. (2012) introduced the concept of an extended gate: a dry port for which the deep-sea terminal can choose to control the flow of containers to and from that inland terminal. The combination of intermodal planning and such a new business model is in recent years referred to as synchromodal transportation (Lucassen et al., 2012, SteadieSeifi et al., 2014, Behdani et al., 2014, Tavasszy et al., 2015). These studies mention the flexible deployment of modes, the possibility of last minute changes to the transportation plan (switching) and a central network orchestrator that offers integrated transport.

In such a synchromodal network, customers of the network operator do not book transports on specified services, but place orders with specific delivery time requirements. The network operator accepts orders without regarding the service schedule, considering some threshold (e.g. a minimum delivery time of 24h).

(16)

Chapter 1 – Introduction 3

Subsequently, the orders are planned on the transportation network, minimising costs and satisfying delivery time requirements as much as possible. In practice multiple problems need to be addressed in order to operate such a synchromodal network. With this dissertation, we aim to propose solutions for three aspects to enable synchromodal networks in practice.

1.1.1 Problem Statement

The main challenge for a container transportation network operator is the continuous construction of an efficient transportation plan. That is, the allocation of containers to available inland services (train, barge or truck). Creating the transportation plan for the network of inland services is referred to as planning in this dissertation, i.e. allocation of all orders to available services in the network. Creating more planning flexibility should help to raise the utilisation rate of inland barge and rail capacity and thus decreasing costs and emissions. Also, the planning flexibility can be used to deal with uncertainties and disturbances, and thus increasing the on-time performance and reliability of the transportation.

In this study, we define a corridor as a direct connection between a deep-sea terminal and an inland terminal area. In practice, the inland transportation in North-West Europe is generally considered per corridor and not for the network as a whole. Based on our experience with practice, this is the case for mainly three reasons: Firstly, no suitable methods for creating an integrated network plan exist yet. Secondly, adapting the plan in real-time responding to delays and other changes occurs manually, by planning operators that focus on specific corridors and inland connections. Thirdly, because of the customer’s restrictions for its transportation orders, the network orchestrator misses the flexibility to switch between modes and routes and thus cannot achieve the benefits of synchromodal planning. In Section 1.3, we will show the research gaps in literature on these topics. The research for this dissertation has been motivated by the development of synchromodal container networks. Our goal is to address the literature gaps and develop methods for optimal portfolios and optimal transportation plans that

enable synchromodal planning in inland container networks. For achieving this

objective, three topics of research are considered in more detail:

1. Integrated network planning: Methods for creating an integrated transportation plan for intermodal transportation networks that are operated by a network orchestrator.

2. Methods for real-time network planning: Methods for creating the transportation plan in real-time and updating it continuously as new information arrives.

3. Balancing customer value and planning flexibility: Methods for optimising a differentiated portfolio to allow flexible transportation planning.

(17)

With this chapter, we do not aim to provide a complete overview of all developments on synchromodal transportation, but merely an overview of the ongoing research for the Rotterdam case from practice. This chapter highlights recent developments and introduces the three topics of research that this dissertation contributes to. All three topics are focused on implementing the concept of synchromodal container transportation: optimisation of integral network planning, methods for real-time decision making for planning and the creation of a product portfolio that allows for more flexibility in the network planning problem.

1.2 _{Case study}

In this dissertation we will investigate to what extent a synchromodal business model contributes to efficiently planning intermodal networks, since without flexibility, little network optimisation is possible. We consider these developments with respect to the case of European Gateway Services (EGS), a subsidiary of the container terminal operator Hutchison Ports ECT Rotterdam (ECT). EGS started with the introduction of regular train services on the corridor between Rotterdam and the inland terminal TCT Venlo. Fig. 1.1 shows the EGS network in 2012, with three deep-sea terminals in Rotterdam and seven hinterland terminals. In 2017, the concept has been extended to 21 hinterland terminals, and a yearly throughput of over 800.000 TEU (European Gateway Services, 2017). The network operator runs over a hundred weekly barge and rail services between the deep-sea ports of Rotterdam and Antwerp and the inland destinations. In four of the inland terminals, ECT has a stake, while the other terminals are third parties. Some of the inland services are fully operated by EGS (with long-term lease of capacity), while other services are carried out in cooperation with other operators, for the purpose of risk and capacity sharing. ECT started with the development of the EGS hinterland network in 2007. Its aim is to strengthen ECT’s position in the European hinterland. Adding network connections is a time-consuming process, since market capture depends largely on frequency, and proven reliability, which both is difficult to achieve on new corridors. As a network operator, EGS takes incoming orders for transportation between a deep-sea location and an inland location (or vice versa). When a container arrives at an inland terminal, the customer can arrange for a pick-up by truck. In some cases, the transportation order is to deliver at an inland location (e.g. a warehouse); in such cases EGS also takes care of the last-mile transportation between an inland terminal and the inland location. The last mile transportation can be carried out by terminal trucking equipment, or by subcontracted third parties.

At the start of our research, in 2012, EGS had no integral planning approach available, yet. Instead, each corridor was planned and operated separately, mostly by accepting orders on a first-come-first-serve basis. Therefore, from an operational point of view, in such a setting the corridors do not support each other, and customers are not provided with alternatives in the case of disruptions. EGS is

(18)

continuously developing on two fronts: on the one hand integrating the planning process for the entire network and on the other hand offering a lead-time based transport portfolio that allows switching of containers. This process has shown to be dependent on several factors, such as information exchange with partners, IT developments and a mind shift of all people involved (sales executives, planners and customer contacts).

The company’s goal is to provide network-wide synchromodal transportation, meaning to optimise all network transportation in an integrally operated network, making use of all transportation options in the most flexible way. A more general description of synchromodal transportation follows in the Section 1.3.1.

Fig. 1.1 Overview of connections in the EGS network in 2012 (EGS,2012)

1.3 Literature overview

In this section, first an overview of literature on the topic of synchromodal transportation is provided. Subsequently, relevant literature on three topics is presented: methods for integrated network plans, methods for real-time planning and methods for creating planning flexibility. We highlight the literature gaps that we will address.

(19)

1.3.1 Synchromodal transportation

In recent years a large amount of literature has been published on the topic of synchromodal transportation. Most studies focus on creating efficient transportation plans, as is the purpose in the long line of research of intermodal planning problems, noted in the overviews of Caris et al. (2013), SteadieSeifi et al. (2014), Reis (2015) and Dong et al. (2017). The recent studies into synchromodal transportation generally aim to include more practical elements into the more general models of intermodal transportation as in Crainic and Kim (2007). These new elements in the models usually depend on the perspective of the researcher and together create an ambiguous definition of the concept of synchromodality. Pfoser et al. (2016) created a framework to identify critical factors in synchromodality. Based on a literature review of several studies relating to the concept, they identified 7 factors related to synchromodality: cooperation, transport planning, intelligent transport systems (ITS), infrastructure, legal framework, mental shift and service offering. This dissertation’s focus is mostly related to service offering (including pricing) and transportation planning. For this, a network operator can employ a business model with lead-time based transportation services, rather than just selling transportation slots. As such, the network operator gains flexibility to optimise utilisations, and operate the network more efficiently. In this section we provide an overview of recent research contributions on these topics. Several studies focused on efficient network planning in a synchromodal setting, i.e. by considering the combination of committed and uncommitted capacity (Ypsilantis, 2016, pp. 47-82; Van Riessen et al. 2015-a), real-time planning (Nabais et

al., 2013; Van Riessen et al. 2015-b, 2016; Van Heeswijk et al., 2016), generating

options (Kapetanis et al., 2016; Mes and Iacob, 2016) or including vessel routing (Fazi

et al., 2015). Other studies have considered the pricing and properties of

transportation services, usually in combination with logistics planning. For instance, Li et al. (2015) designed a pricing scheme based on average costs, rather than actual costs per itinerary and thus allowing a reduction of the standard price due to network efficiencies. Dullaert and Zamparini (2013) study the impact of lead time variability in freight transport. Crevier et al. (2012) compared a pricing strategy for specific itineraries, with a strategy of pricing per transportation request. Bilegan

et al. (2013) introduced a revenue management strategy of accepting or rejecting

bookings on a railway corridor. Similarly, Wang et al. (2016) consider accept-reject decisions for a barge transportation network, including some customers with long term commitments. Table 1.1 provides an overview of planning-related studies and categorises them regarding the perspective of the optimisation problem, the dimensions of flexibility and the considered decisions. Regarding the optimisation perspective, most studies consider the cost minimisation problem of the transportation provider provided a certain available capacity. This is different from the logistics service provider’s perspective, which usually has no invested capacity and therefore can optimise container transports one at a time. Most studies mention

(20)

to some extent three dimensions of flexibility: mode, route and timing. In Table 1.1, we restricted the categorisation to those dimensions that specifically influenced the modelling choices. Finally, we distinguished between 5 types of decisions: the scheduling of transportations, accepting or rejecting bookings, the deployment of (barge or rail) services, the pricing of transportation services or the conditions of the transportation service. From these decision types, the first typically is aimed at the operational level, whereas the other three are typically tactical decisions. Although it has not been published yet, the work presented in Chapter 6 of this dissertation is added to the table for comparison. The contents of the remaining chapters will be outlined in Section 1.5.

From Table 1.1 it can be observed that most studies consider either the perspective of the transportation provider, or the logistics service provider. Also, most studies considered a problem that combined routing and timing – in most cases, the mode is considered implicitly in the definition of the service schedule. Only some considered mode-specific constraints, such as the potential for rerouting with barges (Fazi et al., 2015) or the possibility of transhipments. Finally, almost all studies considered an operational planning problem, for optimal allocation of cargo to an available schedule. In some cases, this was combined with a service schedule design problem. We address three gaps in the literature. Firstly, existing models for integrated network planning lack elements required for synchromodal planning (Section 1.3.2). Secondly, few models are suitable for applying to practice in real-time (Section 1.3.3). Finally, offering synchromodal services in intermodal networks introduces a new problem to optimally balance customer value and planning flexibility (Section 1.3.4).

(21)

Table 1.1 Overview of synchromodal studies and main differentiators

Perspective Flexibility Decision

Operational Tactical Tr ans po rt ati on p ro vi der Lo gi st ic s ser vi ce p ro vi der Shi pp er M ode* R ou te Ti me Tr ans po rt ati on s chedu ling Acc ep t/ rej ec t o rder s Se rvi ce dep lo yme nt Pr odu ct def in iti ons : (c )o nd it io ns / (p )p ri ci ng Bilegan et al. (2013) • ◦ ◦ • ◦ • • • ◦ ◦ Fazi et al. (2015) • ◦ ◦ • • • • ◦ • ◦ Kapetanis et al. (2016) ◦ • • ◦ • • • ◦ ◦ ◦ Li et al. (2015) • ◦ ◦ ◦ • • ◦ ◦ ◦ • (p) Li et al. (2016) • ◦ ◦ • • • • ◦ ◦ ◦

Mes and Iacob (2016) ◦ • ◦ ◦ • • • ◦ ◦ ◦

Nabais et al. (2013) ◦ ◦ ◦ • • • ◦ ◦ ◦

Rivera et al. (2016) ◦ • ◦ • • • • ◦ ◦ ◦

Van Heeswijk, et al. (2016) ◦ • ◦ ◦ • • • ◦ ◦ ◦ Van Riessen et al. (2015-a) & Ch. 2 • ◦ ◦ • • • • ◦ • ◦ Van Riessen et al. (2015-b) & Ch. 3 • ◦ ◦ ◦ • • • ◦ ◦ ◦ Van Riessen et al. (2016) & Ch. 4 • ◦ ◦ ◦ • • • ◦ ◦ ◦ Van Riessen et al. (2017) & Ch. 5 • ◦ ◦ ◦ ◦ • ◦ ◦ ◦ • (c)

Wang et al. (2016) • ◦ ◦ ◦ • • • • • ◦

Ypsilantis (2016, pp. 47-82) • ◦ ◦ • • ◦ ◦ ◦ • • (p)

Chapter 6 • ◦ ◦ ◦ • • ◦ ◦ ◦ • (c)

* Most studies do consider mode to some extent, often as property of a route. In the table, we have marked a study as considering ‘mode’, only if the study specifically considered mode-related aspects.

1.3.2 Integrated Network Planning

The global throughput in container transportation continues to grow and constitutes a growing portion of the global transportation (Drewry Shipping Consultants, 2007). Meanwhile, supply chains get increasingly interconnected and shippers demand higher levels of service, such as short delivery times and reliability (Crainic and Laporte, 1997; Crainic, 2000; Veenstra et al., 2012). The logistic expression for integrated transportation is intermodality. The International Transport Forum defined intermodal transportation as: Multimodal transport of

goods, in one and the same intermodal transport unit by successive modes of transport without handling of the goods themselves when changing modes (UNECE, 2009). The

planning of intermodal transportation requires a network-wide approach (Crainic 2000; Jansen et al. 2004; Crainic and Kim 2007). Consolidation of flows between hubs in intermodal networks is cost efficient as it benefits of the economies of scale

(22)

(Ishfaq and Sox, 2012). Transportation used to be optimised based purely on costs. However, Crainic and Laporte (1997) signal that carriers and transporters cannot only optimise the transportation on cost efficiency anymore. Apart from low tariffs, customers demand for a higher quality of service. According to Crainic and Laporte, quality of service consists of three parts: on-time delivery (time window), delivery speed (service time) and consistency of these aspects. Veenstra et al. (2012) mention reliability as an important quality of service. Ishfaq and Sox (2010) mention six performance targets for intermodal logistic networks: cost, service frequency, service time, delivery reliability, flexibility and safety. They propose methods to optimise the costs of intermodal logistic networks, while meeting service time requirements. The other performance targets are neglected in their work.

Some of the existing tactical service network formulations use strict constraints on delivery time (Ziliaskopoulos and Wardell, 2000) or no due time restrictions (e.g. Crainic, 2000). Strict constraints do not accurately model the flexibility that transportation planners have in consultation with customers. If no time restrictions are considered at all, the existing time pressure in the container transportation is neglected. Several formulations model the economies of scale that occur when cargo is consolidated on an arc (e.g. Ishfaq and Sox, 2012). These abstract formulations of economies of scale cannot directly represent the current situation. The current practice in intermodal container networks is that multiple service and terminal operators cooperate and in this perspective, economies of scale are exploited by selecting services operated by the network operator (self-operated services) or use subcontracted transport. The difference in cost structure between these two cannot be modelled in the existing formulations for the economies of scale. We aim to address these two aspects in the problem of integrated network planning: accurately modelling time pressure, while allowing overdue delivery and modelling a combination of self-operated and subcontracted services. The first issue was also studied by Arikan et al. (2014) for the Danube region between Southern Germany and Hungary. They applied a stochastic service network design model with penalties for overdue deliveries.

1.3.3 Real-Time Network Planning

For efficient synchromodal transport plans it is essential to allow real-time switching, i.e. real-time planning updates. This was recognised by many studies that refer to synchromodal transportation (Lucassen et al., 2012, SteadieSeifi et al., 2014, Behdani et al., 2014, Tavasszy et al., 2015), but not many real-time planning methods that provide a network-wide plan exist yet. The previous section mentioned various planning models that are aimed for solving the network transportation problem offline (Crainic and Laporte, 1997; Crainic, 2000; Crainic and Kim, 2007; Ishfaq and Sox, 2010, 2012, Van Riessen et al. 2015-b). Ziliaskopoulos and Wardell (2000) and Janssen et al. (2004) proposed an online method, but focused on the planning of single corridors. Nabais et al. (2015) and Li (2016) proposed more

(23)

advanced methods for solving the online problem. Their methods uses model predictive control to achieve a required modal split, but the approach requires real-time automated data processing and is less insightful to human planning operators. Li et al. (2013) used a sequential linear programming method. We aim to facilitate synchromodal planning by proposing a real-time network planning solution that is insightful for human planning operators.

1.3.4 Balancing Customer Value and Planning Flexibility

As Van der Horst and De Langen (2008) stated, the container inland transportation chain lacks information integration and stakeholders do not fully trust each other, making integrated solutions difficult. Nonetheless, creating more planning flexibility is vital to enable synchromodal planning. Therefore, the network operator has an incentive to introduce a range of transportation services with varying levels of flexibility. Such new product ranges have been studied recently by Lin (2014) and Wanders (2014). These propositions consider different tariff classes for varying levels of service and the level of decision flexibility that the network operator receives from the customer. In other areas of transportation, incentives of stakeholders are studied with stated preference surveys, e.g. for the valuation of time for travellers (Wee et al., 2013): “travellers are confronted with hypothetical choice situations between a fast, expensive alternative and a cheap one”. To our knowledge, no stated preference studies exist that looked specifically into customer incentives for container transportation, although some studies are in progress (Khakdaman, 2017).

In aviation, the development of revenue management (RM) enabled these industries to increase utilisations (Carmona-Benítez, 2012), e.g. by “selling the right seats to the right customer at the right time” (Zeni, 2001) and by creating customer incentives for using flexible services (Petrick, 2012). The concept of different service propositions in transportation is very similar to the concept of different fare classes for the same flight in aviation. Barnhart et al. (2003) give an overview of operations research in airline revenue management. The primary objective of airline revenue management models is to determine the optimal fare mix: how much seats of each booking class should be available, provided the demand forecasts and the limited total number of seats? Some studies on revenue management in freight transportation focused on an online policy: whether to accept or reject an incoming order. Pak and Dekker (2004) proposed a method for judging sequentially arriving cargo bookings based on expected revenues. If the direct revenue of a booking exceeds the decrease in expected future revenue, the order is accepted. Bilegan et al. (2013) apply a similar approach on rail freight application and Wang et al. (2016) consider accept-reject decisions for barge transportation. In their approach the decision of accepting or rejecting an arriving transport order is based on the difference in expected revenue with and without that order. We propose a solution at a more tactical level, by translating fare mix models from airline revenue

(24)

management towards the setting of intermodal hinterland transportation of containers. The setting of container transportation introduces a new issue to the fare mix problem, as the operator has the opportunity to select from various transport modes, routes and time for some of the containers.

1.4 Research questions and approach

As introduced in Section 1.1.1, the main goal of this dissertation is to address the literature gaps and develop methods for optimal portfolios and optimal transportation plans that enable synchromodal planning in inland container networks. In this section our main and supporting research questions are provided as well as our research approach.

1.4.1 Main research question

In order to enable synchromodal planning in inland container networks in practice, our main research question is as follows:

How can synchromodal networks operate optimally?

Our research question aims for practical solutions, since it focuses on network operation. At the same time, it aims for finding optimal methods for different aspects of network operation. We consider this main research within the scope of synchromodality. As introduced in Section 1.3.1, synchromodality has an ambiguous definition, based on various publications (Table 1.1). In this dissertation, we consider synchromodality as the combination of a service-based business model and network-wide intermodal operation. We consider the perspective of the network operator. Our aim is to maximise the profits of the service-based business model and minimise costs of the network-wide intermodal operation, within acceptable service levels. Based on the literature overview of Section 1.3, our research must address three subtopics in order to answer the main research question: methods for integrated network plans, methods for real-time planning and methods for creating planning flexibility. All these three aspects contribute to the development of synchromodal transportation to such a level that it can be implemented in practice. Sections 1.4.2-1.4.4 introduce the sub research questions regarding these topics and provide an overview of our approach. Finally, all topics combined must lead to a synchromodal network than can be operated and monitored optimally in real-time.

(25)

1.4.2 Integrated Network Planning

The first aspect of our study considers the development of planning in integrated networks. Container transportation is currently organised with A-B connections. However, a network operator carries out services to several closely located inland terminals in the hinterland. A service network between all network locations provides more alternative routes using intermediate transfers. This allows consolidation of flows and an increase of overall capacity. Existing service network design methods are not applied in practice for several reasons: models with more flexible time restrictions are required and self-operated and subcontracted services must be combined. The following research questions are studied:

1. How must a service network design model accommodate for flexibility in overdue delivery as well as subcontracted and self-operated services? In this research an exact method is developed to determine the optimal number of services on all corridors in the network. The service network design must incorporate combinations of self-operated and subcontracted transport and allow for overdue delivery (at a penalty cost) to model current container transportation networks.

Besides, the online planning of the network transportation is important, dealing with continuous disturbances in the network. In case of disturbances, the manual planners have to switch disturbed containers to other routes. This is time-consuming and the network potential for alternatives is often not fully used in practice. Last-minute switching is difficult, resulting in delays. For this, an assessment of the impact of disturbances must be developed. With this assessment, the network operator can find the most important network aspects to improve for increasing reliability and robustness of the transportation and decrease the cost impact of disturbances. The following research questions are considered:

2. How can optimal transportation plans be created for synchromodal networks?

3. How can the effect of disturbances in synchromodal networks be quantified?

With our proposed method, we aim to compare the quality of online updates of an automated optimal method and a method that mimics the manual updates for various disturbances. This provides insight in the gravity of disturbances and the benefit of automating online planning updates. With this part of the research the new Linear Container Allocation model with Time-restrictions (LCAT) is developed, which will be the bases of the research questions for real-time network planning in the next section.

(26)

1.4.3 Towards Real-Time Network Planning

The second research topic aims for enabling real-time network planning, to allow synchromodal transportation in practice. As mentioned in Section 1.3, several studies have proposed optimisation methods for determining the optimal allocation of containers to all available inland transportation services, considering capacity, costs, lead times and emissions. The proposed methods are suitable for solving the

offline planning problem, in which an optimal network plan is created for a batch of

transportation orders collectively. From practice we found three issues with the implementation of a centralised offline approach in intermodal networks:

 The nature of the inland transport logistics requires a real-time approach, and does not allow for integral planning models that are applied in intervals.

 Proposed centralised optimisation methods depend strongly on automation, both for terminals, as for other parts of the supply chain. Such an automation level is often not easy to implement. On top of that, information from direct communication between manual operators is often essential (Douma, 2008).

 Finally, the supply chain of container logistics lacks information integration (van der Horst and de Langen, 2008). In the case of intermodal networks, manual planning operators often do not have real-time capacity information about the inland services.

We aim to find a solution for the following research question:

4. How can the results of the LCAT model be translated into a white box decision support method for human planning operators?

To answer these questions, a general method for obtaining a real-time decision support system (DSS) is required that addresses all three aforementioned issues. The model must be based on an intrinsic analysis of the offline LCAT model, translating the offline model’s optimal solutions to a decision tree for online decision support. A decision tree is a white box method that is comprehensible for manual planners and allows manual changes if necessary. It will therefore more easily be accepted for use in daily practice. The human planner responsible for a central network planning must be able to check available capacity on a proposed service manually. Hence, real-time up-to-date information is not critical for the method’s performance. Note that this method aims to support planning decisions for incoming transportation orders, however, the effect of real-time decision support in case of disturbances or disruptions is not included in the study. In case of a disruption during the operational phase, a different type of real-time decisions must be made in order to solve the disruption and fulfil all transportation requests.

(27)

1.4.4 Balancing Customer Value and Planning Flexibility

The research topics on integral and real-time network planning as described in previous sections provide insights into efficient network planning, from an offline and a real-time perspective, respectively. Also, we highlighted the potential gain in network performance with more planning flexibility. However, in the prevalent set-up of the transportation product, customers are hesitant to transfer planning flexibility to the (network) operator. This is for several reasons, i.e. company policy, habituation, but also the pricing mechanism. Achieving planning flexibility requires persuading clients to allow flexible planning of their transportation orders. For that reason, studies into creating planning flexibility are required to answer the following research questions:

5. What is the value of planning flexibility for synchromodal networks? As suggested by Lin (2014) and Wanders (2014), the market for inland container transportation can be segmented in groups of customers with different characteristics. These groups are sensitive to different incentives that may persuade customers to allow flexibility for planning purposes by the synchromodal transportation operator. For this, a revenue management (RM) model for container logistics is necessary, in order to balance customer demand and network transportation options, similar to revenue management problems in aviation (Barnhart et al., 2003). Currently, only qualitative studies into customer preferences have been carried out for container transportation in North-West Europe, e.g. Lucassen et al. (2010), Palmer, et al. (2012) and Veenstra et al. (2012). One issue with developing a RM model in practice is the high number of stakeholders involved in a container transport. The decision on service level and price often has to be made between several stakeholders with conflicting incentives, such as the cargo owner, the container owner (shipping line) and the logistic service provider (Van der Horst and De Langen, 2008). Therefore, research on various topics is required. First of all, market research is necessary to gain insights in the incentives of different segments of transportation customers. Secondly, using the information from such market research, a method must be created for designing transportation products that encourage flexibility and thus synchromodal transportation. By designing transportation products properties according to customer preferences, customers can be targeted with different types of service level (delivery time, reliability), availability and other aspects. This allows addressing service needs more specifically, and enables pricing mechanisms that maximise revenue, by differential pricing (Barnhart et al., 2003). Thirdly, a pricing strategy must be developed. Currently, transportation is priced per service, based on the mode (barge, rail) and the distance. This is typically cost-plus pricing. If the network operator is using a network planning approach to allocate containers to different modes or routes, this pricing mechanism is not suitable: a customer is not willing to pay a high price if his container is planned on an expensive route for the benefit of the entire network plan. The new pricing strategy must balance the need for flexibility in the order pool

(28)

with maximising revenue. For instance, orders that allow flexible routing with flexible modes may incur a discount on the price. With a value-based pricing model, the price is based on the customer value of a transportation service. In this way a differentiated portfolio can be created, aimed for various customers groups that are willing to pay a certain price for a transportation product with a certain amount of flexibility.

Finally, such a new revenue management strategy in synchromodal transportation networks is different from other applications, such as aviation. With synchromodal networks, the network operator can use the flexibility in some products to attain a more efficient transportation plan. In this case the pricing strategy is strongly linked to the operations management: promoting planning flexibility is beneficial for the network if the flexibility can be used to achieve a more cost-efficient transportation plan. This is depicted in Fig. 1.2. While the operations management aims to assign transportation slots to a provided set of demand for minimum cost, the revenue management strategy aims to attain demand for a provided set of slots with maximum revenue. In our case, these two approaches are connected by the balance between flexibility and network utilisation. To optimise total profit, these two approaches must be optimised integrally to answer the following research questions:

6. How can the optimal fare class mix for a synchromodal corridor be found? 7. To what extent is it relevant to consider the synchromodal network

structure when optimising the fare class mix?

For this new type of problem, we provide a framework for finding the optimal transportation service portfolio, the Cargo Fare Class Mix problem. Our proposed framework is based on the outcome of original market research, product design and a pricing strategy developed by EGS. The CFCM problem considers two transportation services with differentiated delivery lead times. We propose a method to find the revenue maximising balance between those two products. By this integral analysis of revenue and operations management, we aim to shown the value of planning flexibility in synchromodal container transportation. Furthermore, we propose an improved (faster) method for the single corridor problem, and use this to find lower and upper bounds for specific types of synchromodal networks.

(29)

Fig. 1.2 Revenue management and operations management

1.5 Outline of the dissertation

In the next chapters, our research on these three topics is provided. Here, an outline of the remainder of this dissertation is provided. A schematic overview is provided in Fig. 1.3. Currently, the most important developments of synchromodal transportation occur in The Netherlands and focus on the Rotterdam hinterland. In this chapter we have provided a general overview of recent and current developments on the topic of operational implementation of synchromodal transportation. To support this transformation, our research focuses on three topics:

 Models for integrated network planning. In Chapter 2, the tactical level is assessed, for which we developed a new service network design method,

answering research question 1. Chapter 3 considers the impact and relevance of disturbances in a synchromodal container transportation network on the

operational level, and introduces the newly developed LCAT model, answering research question 2 and 3.

 Methods for real-time decision making for network transportation planning. In Chapter 4 a real-time decision support system is described, based

on the LCAT model of Chapter 3. This chapter answers research question 4.

 Methods for optimising a differentiated portfolio of transportation services. In Chapter 5, a solution for finding the optimum in a single corridor case is provided by introducing the Cargo Fare Class Mix (CFCM) problem,

answering research question 5 and 6. Finally, Chapter 6 extends the method towards synchromodal networks, answering research question 7.

The assumptions in the analyses differ between chapters. Since Chapter 2 and 3 mostly focus on the integrated planning of synchromodal networks, the planning conditions are modelled in more detail, whereas the specific customer demands are

Demand (trans-portation orders) Supply (available transporta-tion slots) Operations management Revenue management Flexibility Utilisations

(30)

not in detail considered. In Chapter 5 and 6, the optimal combination of products in the portfolio is considered; therefore, these chapters put more emphasis on customer requirements, and less on planning conditions. Chapter 4 functions as a bridge between planning and practice, aiming to translate model information to real-time support in practice.

The results of our research are expected to have practical relevance, as all case studies are based on the network of EGS and the Rotterdam hinterland. Chapter 7 provides the overall conclusions and answer to the main research question. It also includes suggestions for future research and a description of the practical impacts of our research for EGS hitherto.

Fig. 1.3 Overview of the strucure of the research

Ch. 2. Service network design Ch. 4. Online planning by decision rules Ch. 3. Impact of disturbances & LCAT model Section 1.3.1. Integrated network planning Section 1.3.3. Balancing value and flexibility Network operations management

Section 1.3.2. Towards real-time network planning

Network revenue management

Ch. 6. Cargo fare class mix for synchromodal

networks

Ch. 7. Enabling synchromodal planning in inland container networks: overview and implications

Ch. 5. Cargo fare class mix for

intermodal corridors

(31)

(32)

2 Service network design for an intermodal

container network

In the previous chapter, the concept of synchromodal transportation was introduced, as well as the case of EGS. To use this network cost-efficiently, a centralised planning of the container transportation is required, to be operated by the deep-sea terminal. In this chapter, a new mathematical model is proposed to determine the optimal service schedule between the given network terminals. The model introduces two new features to the intermodal network-planning problem. Firstly, overdue deliveries are penalised instead of prohibited. Secondly, the model combines self-operated and subcontracted slots. The model considers self-operated or subcontracted barge and rail services as well as transport by truck. In this chapter, we provide the answer to research question 1: “How must a service network design model accommodate for flexibility in overdue delivery as well as subcontracted and self-operated services?” In a case study of the EGS network, the benefit of using container transportation with intermediate transfers is studied. The results indicate that the proposed model is suitable for the service network design in modern intermodal container transport networks. Also, the results suggest that a synchromodal business model for the network transport and terminals is worth investigating further, as the transit costs can be reduced with lower transfer costs. This chapter is organised as follows. Section 2.1 provides an introduction to the problem addressed. Section 2.2 briefly reviews literature on service network design models. Section 2.3 introduces the proposed intermodal container network model. The case of EGS is used as an example for the intermodal container network model of this study in Section 2.4. The results of the experiments are discussed in Section 2.5. Section 2.6 concludes the chapter and proposes further research.2

Keywords: Service network design; container logistics; intermodal transportation; hinterland transportation; flexible due times; subcontracted transport

2_{This chapter is based on the following publication with small modifications: Van Riessen,} B., Negenborn, R. R., Dekker, R. and Lodewijks, G. (2015). Service network design for an intermodal container network with flexible transit times and the possibility of using subcontracted transport. International Journal of Shipping and Transport Logistics, 7(4), 457-478. The final publication is available via https://doi.org/10.1504/IJSTL.2015.069683.

(33)

2.1 Introduction

2.1.1 Development of container networks

In Section 1.1 an overview of recent developments in container networks was presented. A tendency of more integrated supply chains has sparked initiatives in North-West Europe to create transportation networks for containers (Groothedde

et al., 2005, Lucassen and Dogger, 2012, Rodrigue and Notteboom, 2012, Port of

Rotterdam, 2011). In Fig. 1.1 the network of European Gateway Services (EGS) is depicted, a subsidiary of Europe Container Terminals (ECT), with three deep-sea terminals in Rotterdam and seven hinterland terminals. In this chapter, the service network design of such container networks is considered. For organising the transportation, a network operator such as EGS does not own barges and trains. It uses a combination of long-term contracts for a fixed amount of services per week (self-operated services) or it uses slots on existing services on a per-slot-basis (subcontracted transport). A long term contract comes with the risk of not fully loading the available capacity that is already paid for, but it also brings economies of scale. For subcontracted transport the network operator incurs no risk of unutilised space, but this comes at a higher price. For both types, a transport operator carries out the actual transport. Based on these observations, we propose a service network design model with several new aspects. First, the next section provides relevant definitions.

2.1.2 Definitions: intermodal and synchromodal

This study focuses on the transportation from the seaport terminal to a hinterland terminal (import) or vice versa (export), organised by the deep-sea terminal. This is called hinterland transportation. Final drayage to a customer is excluded. In the network, transport is carried out by three different modes: barge, rail and truck. Hence, as different modes can be selected, the transportation in the network is considered multimodal transportation. At terminals, containers can be switched from one transport mode to another. In this chapter, an exchange at a terminal is called a

transfer. Fig. 2.1 shows a schematic view on three terminals. The figure shows five

mode-specific corridors by which the terminals are directly connected. As multiple modes connect two terminals, multiple corridors exist. Terminal A and C are indirectly connected via terminal B, and transport is possible using the corridors to B and then to C. Each of the transport steps from one terminal to another is called a

leg. The two consecutive legs are referred to as a connection between A and C. The

specific itinerary of a container, i.e. the services used, is called a path. Each of the used corridors is referred to as a leg of the container transport. The service on a corridor between terminals is the movement of a vehicle from one terminal to another, travelling on a specific time and route. The number of services per time

Optimal Transportation Plans and Portfolios for Synchromodal Container Networks

ERIM PhD Series

Research in Management

Optimal Transportation

Plans and Portfolios for

Synchromodal Container

Networks

BART VAN RIESSEN

Optimal Transportation Plans and Portfolios for

Synchromodal Container Networks

Bart van Riessen

Erasmus University Rotterdam

Delft University of Technology

Optimal Transportation Plans and Portfolios for

Synchromodal Container Networks

Optimale transportplanning en portfolio’s voor

synchromodale containernetwerken

Acknowledgements

Table of Contents

Clarification of contribution

1

Introduction

1.1

Background and problem statement

1.2

Case study

1.3

Literature overview

1.4

Research questions and approach

1.5

Outline of the dissertation

2

Service network design for an intermodal

container network

2.1

Introduction

_{Case study}